[00:45] <tsimonq2> so the Kernel newsletter says that the current cycle is a cycle that ended three days ago, and the next cycle has already begun...what's up with that?
[00:45] <tsimonq2> https://wiki.ubuntu.com/KernelTeam/Newsletter/2016-02-23 for reference
[08:21] <marlinc> What would be the recommended channel to ask questions on driver related stuff. In this case its about a Iiyama touchscreen
[08:39] <apw> tsimonq2, that'd be an error by the sounds of it
[08:40] <apw> tsimonq2, it is easy to forget to update that, as the newsletter is weekly, but the cycles change 3 weekly
[08:40] <apw> jsalisbury, ^
[08:41] <apw> marlinc, you could ask here, we may redirect you after we find out what the nature of the question is
[08:44] <marlinc> I've got a Iiyama ProLite T2452MTS touchscreen but its not registering anything
[08:45] <marlinc> It does show up as a input device
[08:45] <marlinc> Listening to it with 'input-events' doesn
[08:45] <marlinc> Listening to it with 'input-events' doesn't show anything though so I'm not sure how to continue from there
[08:45] <marlinc> It shows up as 'Bus 003 Device 014: ID 093a:8002 Pixart Imaging, Inc.'
[08:52] <apw> marlinc, which series/kernel are you running ?
[08:52] <marlinc> Description:	Ubuntu Xenial Xerus (development branch)
[08:52] <marlinc> Linux marlinc-laptop 4.4.0-6-generic #21-Ubuntu SMP Tue Feb 16 20:32:27 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
[08:59] <smb> marlinc, in theory this site states it should be using the generic hid-multitouch. Though it might not have an alias to load it... http://lii-enac.fr/en/architecture/linux-input/multitouch-devices.html
[09:03] <marlinc> How would I try that out smb, not sure if I currently have the wrong module loaded
[09:04] <smb> hm, no... I can see that id (though as hid) in the hid-multitouch driver even in Trusty... question would be whether the usb devide is bound to somehting... think that can be seen via sysfs
[09:05] <marlinc> https://gist.github.com/Marlinc/7536ba44d19287233824
[09:06] <marlinc> What would I have to do to check that smb 
[09:07] <smb> marlinc, cannot find it again but your output shows that it got detected by the hid driver at least
[09:07] <marlinc> Stuff in /sys/bus/hid/devices/0003\:093A\:8002.0011/ ?
[09:07] <smb> so ls-input should list it
[09:07] <marlinc> It does list it, wait
[09:08] <smb> and then input-events 0 
[09:08] <smb> should print events
[09:08] <marlinc> The issue is that it doesn't show anything
[09:08] <marlinc> https://gist.github.com/Marlinc/6c23677ccd053bf851b5
[09:08] <smb> oh wait
[09:08] <marlinc> Tried that already ;)
[09:08] <smb> no input45?
[09:09] <marlinc> https://gist.github.com/Marlinc/c6ce012dd2dccfbaf6c8
[09:09] <marlinc> That's it
[09:12] <smb> marlinc, seems it shows up there as event11
[09:12] <smb> So input-events 11
[09:14] <marlinc> Yea, that's the link above that. Doesn't show anything when I touch the screen
[09:17] <smb> Hm, that would be as much as I could help. Might be broken, might be the vendor having replaced the complete internals while keeping the old id...
[09:18] <marlinc> Probably not going to work then :p
[09:19] <marlinc> Lets see if I can find a manual that hopefully has information on the support
[09:19] <marlinc> Might have to enable it, not sure
[09:20] <marlinc> Ah, I guess it does work. You appear to need a 'touchpen'
[09:21] <marlinc> Which doesn't do anything
[09:23] <smb> That sounds a bit backwards... with everything being touch with ones fingers nowadays... 
[09:24] <marlinc> Mm should support finger touch as well..
[09:24] <marlinc> Then I guess its broken
[09:26] <smb> marlinc, Could be. Though I would not say I am absolutely sure there. 
[09:27] <marlinc> Yea, I guess there's no way to test that so :p
[09:27] <marlinc> Wait there is, I can at least see if there's 'traffic' going over the USB connection with Wireshark
[09:30] <marlinc> Interesting, its trying to access the stream from the wrong path. Its trying to open it from /sys/kernel/debug/usbmon/0t but its in /sys/kernel/debug/usb/usbmon/0t
[09:30] <marlinc> I can't create symlinks in /sys so that's a bit problematic :P
[09:36] <marlinc> smb, ^?
[09:39] <smb> marlinc, erm, that would be wireshark using the path, wouldn't it? 
[09:39] <marlinc> tcpdump has the same issue
[09:39] <marlinc> https://bugs.launchpad.net/ubuntu/+source/libpcap/+bug/523349
[09:39] <marlinc> This is the exact issue
[09:39] <marlinc> That patch would fix it even
[09:40] <marlinc> Not sure why the bug is back
[09:44] <smb> marlinc, that could be that the wireshark package got rebased and accidentally dropped a fix. I have not looked there though. At least the kernel path did not change that recently its the usb/usbmon in  3.13 already
[10:05] <marlinc> Do I have to re-open the issue smb?
[10:07] <smb> marlinc, Generally its better to report it as a new bug (even if then someone marks it duplicate). But imo its a new issue to have the wrong path again. In theory you should be able to use "ubuntu-bug wireshark"
[10:08] <marlinc> It doesn't appear to be an issue in wireshark. Its an issue in libpcap. tcpdump has the same issue
[10:08] <genkgo> hello, i read everywhere that linux kernel 4 can be updated without reboot. i am on trusty (14.04.3) with wily kernels. now i upgraded from 4.2.0-27 to 4.2.0-30. but now uname -a still says 4.2.0-27. how can i update without the reboot?
[10:08] <smb> ah ok... so just to replace wireshark with pcap I guess... hope...
[10:09] <smb> I mean libpcap 
[10:09] <smb> whatever the package name is
[10:28] <smb> genkgo, not at all with 4.2. You mean life-patching and that is not supported at all before 4.4. And even with it being possible, it does not mean you can update complete kernels without reboot. There are technical limitations to what can be done without reboot. And even what can be done needs to be provided in a special format.
[10:29] <genkgo> smb: ok, thanks for the explanation. so basically, i just need to reboot, right?
[10:29] <genkgo> for the update from 4.2.0-27 to 4.2.0-30?
[10:29] <smb> marlinc, One other thing you might try is to try a newer mainline kernel (and personally I would enable grub menu to do that)
[10:29] <smb> genkgo, yes
[10:29] <genkgo> alright, thanks
[10:32] <smb> marlinc, Just happened to see some enlantech related stable patch (pulled back from 4.5) on some mailing list. Could be totally unrelated but before having to talk to some support hotline...
[10:47] <marlinc> I could try a new kernel smb 
[10:47] <marlinc> I have ZFS as root fs so I can simply clone the existing FS and play with that
[10:48] <marlinc> Yesterday I played with unity8 that way
[10:52] <marlinc> How would I go about doing that smb?
[10:55] <genkgo> smb: i have another question. i have three ubuntu machines. two are running linux-image-4.2.0-30-generic, but one is running linux-image-4.2.0-31-generic. how come that that machine picks up 31 while the others do not?
[10:55] <smb> marlinc, Hm, rather carefully in that case. If you download and install one of http://kernel.ubuntu.com/~kernel-ppa/mainline/?C=N;O=D you have to be careful to check whether the zfs dkms module still builds. Otherwise there will be panic (first the kernel and then probably you ;)) Thats why I suggest changing grub to remove the hidden timeout
[10:56] <smb> genkgo, Could be that you enabled proposed on that one machine in the past
[10:56] <genkgo> smb, that could be, how can i look that up?
[10:56] <marlinc> Well I guess I have to manually compile ZFS in that case as DKMS is no longer being used in 16.04 for ZFS
[10:57] <genkgo> smb: "deb http://ppa.launchpad.net/canonical-kernel-team/ppa/ubuntu trusty main"
[10:57] <smb> genkgo, either via gui software sources (I think) or grep proposed /etc/apt/sources.list
[10:57] <genkgo> smb does not pick up proposed?
[10:57] <genkgo> does that pick up proposed?
[10:57] <smb> marlinc, Oh right. forgot that got switched
[10:58] <smb> genkgo, actually that picks up packages as soon as they get uploaded... not sure that should be there by default
[11:00] <smb> genkgo, kernels get uploaded and built there and then copied to the proposed pocket. So I would disable the ppa and just enable the proposed pocket (if you want to have kernels as soon as they get into testing and bug verification)
[11:03] <smb> marlinc, Might be much less painful if you had some other box using a more traditional fs to attach to that touchscreen for experiments...
[11:03] <genkgo> smb, thanks, i will remove the ppa
[11:04] <genkgo> smb, i was involved in bug verification, so i think i left it there
[11:07] <smb> genkgo, Hm... maybe at some point in the past that was done that way. But nowadays I would say there is absolutely no need (rather being dangerous) to have the ppa enabled.
[11:10] <genkgo> smb: it was, back then i need to upgrade to vivid kernels. nowadays i can install -vivid and -wily.
[12:40] <tsimonq2> apw: alright thanks :)
[15:27] <dannf> ogasawara: i think i have a pull request stuck in moderation - would you mind taking a look? also - i'm curious about why mailman hates about my mail - let me know if its something i can change on my end.
[15:40] <manjo> apw, anything else you need from me on https://bugs.launchpad.net/ubuntu/+source/initramfs-tools/+bug/1548120/ ? 
[16:07] <ogasawara> dannf: heya, sorry was on a call.  I just checked the queue and I didn't see anything from you.  What was the subject line?
[16:08] <ogasawara> dannf: maybe someone beat me to moderating it
[16:09] <apw> manjo, nope, all uploaded, waiting on the grinders to grind
[16:09] <dannf> ogasawara: np. weird...[PULL][Xenial] Fix ARM KVM guest hangs w/ host ntp
[16:09] <apw> dannf, i think i moderated that before today, odd
[16:09] <manjo> apw, thanks a ton .. the bug is still marked inprogress and block-proposed I think ... 
[16:09] <apw> manjo, don't worry
[16:09] <manjo> ok ☺ 
[16:10] <manjo> apw, cust is watching that bug ... so brought it up 
[16:10] <dannf> apw: oh weird - ok. i just didn't see it come in via the list. /me checks archive
[16:10] <dannf> apw: curious - what is causing my msgs to get moderated?
[16:11] <dannf> apw: yep - msg is in the archive. sorry for the false alarm, i'll blame gmail :)
[16:12] <apw> you must be subscribed to the list under a different alias to you are sending, like for me apw@ or andy.foo@
[16:12] <ogasawara> dannf: indeed, I see you subscribed as dannf@ubuntu.com
[16:13] <dannf> ogasawara: ah, i can fix that.
[20:39] <lamont> jsalisbury: all yours, sir.  I threw some additional info at the bug, too.
[20:40] <jsalisbury> lamont, great, thanks
[20:40] <cristian_c> jsalisbury: hi
[20:40] <jsalisbury> cristian_c, hello
[20:40] <lamont> jsalisbury: presumably there should be no issue with me just living on this kernel, true?
[20:41] <jsalisbury> lamont, ther shouldn't be.  It's the stock kernel with a single revert.
[20:41] <lamont> woot
[20:42] <lamont> I shall test it with passion then
[20:42] <lamont> (that is to say, I
[20:42] <lamont> ll keep using it)
[20:42] <jsalisbury> lamont, great.  I'll contact the upstream patch author for feedback
[20:42] <lamont> it's entirely possible that his delay is simply too short
[20:43] <lamont> actually.... wanna unrevert that, make the delay 3 seconds, and toss me the resulting kernel?
[20:43] <lamont> because 2 orders of magnitude should do the trick, right?
[20:44] <lamont> I almost tried building a kernel in that bisect, but there was no debian/ directory, and I decided that life was short
[20:44] <jsalisbury> heh, yeah, I'll build you another kernel
[20:44] <lamont> I'm totally going off of memory on the commit loog talking about a  30 (or 50) mSec delay
[20:51] <jsalisbury> lamont, looking at the patch, it does 3 retries at 10ms each for a total of 30ms
[20:51] <cristian_c> jsalisbury: have you taken a look at my patch?
[20:51] <lamont> jsalisbury: make it 300 retries and log how many after it passes?
[20:51] <lamont> :D
[20:51] <lamont> this is a big rock, I admit
[20:51] <jsalisbury> cristian_c, I have not yet.  I can submit it upstream for you if you want, or post directions in the bug
[20:52] <cristian_c> jsalisbury: what do you suggest?
[20:52] <jsalisbury> lamont, worth a try.  I'll also ping upstream and see what they think.
[20:52] <jsalisbury> cristian_c, either way is fine.  I can try to submit it this week.
[20:52] <lamont> jsalisbury: definitely logging the count, because I expect that to be hilarious
[20:53] <cristian_c> jsalisbury: ok